Method for operating a radar system that determines the location of a detected object using machine learning, radar system, driver assistance system, and vehicle

Abstract

A method of operating a radar system (14), particularly a radar system (14) for a vehicle (10), a radar system (14), a driver assistance system (12), and a vehicle (10) is described, in which at least one electromagnetic radar signal (22) is transmitted from at least one transmit antenna element (34) of the radar system (14). At least one electromagnetic echo signal (40) resulting from the at least one radar signal (22) reflected from at least one object (20) within the field of view of the radar system (14) is received by at least one receive antenna element (38) and converted into receive data (72) suitable for signal processing. The radar system (14) operates as a MIMO radar with multiple transmit antenna elements (34) and multiple receive antenna elements (38), and during MIMO operation of the radar system (14), the transmit antenna elements (34) and the receive antenna elements (38) generate a virtual antenna array (42) having multiple virtual antenna elements (44) for receiving echo signals (40). Magnitude information and phase information are each determined from at least a portion of the received data by performing at least one two-dimensional fast Fourier transform on at least a portion of the virtual antenna elements. At least one array data set is determined, including at least the magnitude information and the phase information for at least a portion of the virtual antenna elements. At least a portion of the data in the at least one array data set is provided to at least one neural network (64), thereby determining at least one location information of at least one detected object (20).

Description

[Technical Field]

[0001] The present invention relates to a method of operating a radar system, in particular a radar system for a vehicle, comprising the steps of: at least one electromagnetic radar signal is transmitted from at least one transmit antenna element of the radar system; At least one electromagnetic echo signal resulting from at least one radar signal reflected from at least one object within the field of view of the radar system is received by at least one receive antenna element and converted into received data suitable for signal processing; at least one magnitude information and at least one phase information corresponding to at least one received echo signal is determined from at least a portion of the received data; At least one piece of position information is determined using machine learning, and the at least one piece of position information characterizes at least the orientation of at least one detected object relative to a frame of reference for the radar system.

[0002] Furthermore, the present invention provides at least one transmitting antenna element for transmitting an electromagnetic radar signal; at least one receive antenna element for receiving electromagnetic echo signals; means for converting the electromagnetic echo signals into received data suitable for signal processing; means for determining magnitude and phase information from the received data; means for determining, using machine learning, at least one positional information characterizing at least the orientation of at least one detected object; The present invention relates to a radar system, in particular an automotive radar system, comprising:

[0003] Furthermore, the present invention relates to a driver assistance system comprising at least one radar system, wherein said at least one radar system comprises: at least one transmitting antenna element for transmitting an electromagnetic radar signal; at least one receive antenna element for receiving electromagnetic echo signals; means for converting the electromagnetic echo signals into received data suitable for signal processing; means for determining magnitude and phase information from the received data; means for determining, using machine learning, at least one positional information characterizing at least the orientation of at least one detected object; The present invention relates to a driver assistance system comprising:

[0004] Still further, the present invention provides a vehicle comprising at least one radar system, the at least one radar system comprising: at least one transmitting antenna element for transmitting an electromagnetic radar signal; at least one receive antenna element for receiving electromagnetic echo signals; means for converting the electromagnetic echo signals into received data suitable for signal processing; means for determining magnitude and phase information from the received data; means for determining, using machine learning, at least one positional information characterizing at least the orientation of at least one detected object; The present invention relates to a vehicle equipped with the above.

[0005] U.S. Patent Application Publication No. 20210156985A1 describes techniques and apparatus for resolving radar angle ambiguity. These techniques enable the determination of a target's angular position from a spatial response with multiple amplitude peaks. Rather than simply considering the highest amplitude peak, the radar angle ambiguity resolution technique selects a frequency subspectrum (or multiple frequency subspectrums) that emphasizes amplitude or phase differences in the spatial response and analyzes the irregular shape of the spatial response across a wide field of view to determine the target's angular position. This results in a unique signature for each target's angular position, which the radar system can determine to resolve the angular ambiguity. Using these techniques, radars can achieve antenna array element spacing greater than half the central wavelength of the reflected radar signal used for target detection. The radar system determines the target's angular position by detecting the radar signal reflected from the target and identifying the steering angle corresponding to the radar signal's angle of arrival. Digital beamforming is used to generate a spatial response containing amplitude and phase information for different steering angles. An angle estimator receives the spatial response and estimates the target's angular position by analyzing the shape of the spatial response across the field of view. In some implementations, the angle estimator may use signal processing techniques, pattern matching techniques, or machine learning to determine the angle of arrival of the radar signal. Exemplary signal processing techniques may utilize algorithms that analyze the shape of the spatial response and determine differences that indicate the direction of the target.

[0006] It is an object of the present invention to provide a radar system, a driver assistance system and a vehicle that can improve the determination of position information regarding objects detected by the radar system, in particular the direction of the objects. Summary of the Invention

[0007] The object of the present invention is to operating the radar system as a multiple-input multiple-output (MIMO) radar including a plurality of transmit antenna elements and a plurality of receive antenna elements, wherein during MIMO operation of the radar system, the transmit antenna elements and the receive antenna elements form a virtual antenna array having a plurality of virtual antenna elements for receiving echo signals; magnitude information and phase information are each determined from at least a portion of the received data by performing at least one two-dimensional fast Fourier transform on at least a portion of the virtual antenna elements; at least one array data set including at least the magnitude information and the phase information of at least the portion of the virtual antenna elements is determined; at least a portion of the data of at least one of the array data sets is fed to at least one neural network, thereby determining at least one location information of at least one of the detected objects; This is achieved by the method.

[0008] According to the present invention, the radar system is operated as a MIMO (multiple-input multiple-output) radar. A MIMO radar includes multiple transmit antenna elements and multiple receive antenna elements. During MIMO operation of the radar system, the transmit antenna elements and receive antenna elements form a virtual antenna array with multiple virtual antenna elements. In this way, the number of receive antenna elements that serve as sampling points for echo signals can be increased. This also allows for an improved resolution of location information.

[0009] The received data is transformed into at least one array data set by a two-dimensional fast Fourier transform (2D FFT). The at least one array data set includes at least intensity and phase information obtained by at least some of the virtual antenna elements. In this way, the received data is transformed into the intensity / phase domain. The array data set characterizes a 2D FFT spectrum. A characteristic pattern of side lobes is observed in the 2D FFT spectrum of the ambiguity function in the virtual antenna array. This aperture pattern can be referred to as an "aperture pattern." The aperture pattern is relatively simple when only one or two targets are detected, but becomes more complex when there are many targets. The aperture pattern is determined by both the intensity and phase information of each target, as the intensity and phase information are mixed (superimposed) with each other.

[0010] By complexly mixing the received data from each signal of two or more targets within a single range / Doppler cell, unique aperture patterns are formed. By identifying these unique aperture patterns, the location information, particularly the respective directions, of the two or more targets can be separated. Resolution is characterized by the ability to separate two or more targets within the same range / Doppler bin. This resolution is particularly related to the placement of antenna subarrays of the transmit and receive antenna arrays. According to the present invention, multiple targets can be resolved by using ambiguity.

[0011] At least a portion of the data in the array data set is fed to at least one neural network. The at least one neural network determines at least one positional information relating to the position of at least one detected object. By using the at least one neural network according to the present invention, the aperture can be highly sampled to simplify sidelobe patterns. Furthermore, symmetry planes that may result in mirror ambiguity can be resolved.

[0012] Advantageously, for any set of objects and the phase / intensity relationships between them, the aperture pattern as a function of intensity and phase information can be determined. By training at least one neural network to learn which data in the array data sets form each object combination, the location, particularly the orientation, of at least one object can be isolated. By using appropriate simulation techniques, even large initial training array data sets can be systematically generated.

[0013] According to the invention, the intensity information and the phase information, and if applicable the time information, are treated as a single image from which at least one positional information, in particular directional information, can be learned using at least one neural network. The image of the intensity information and the image of the phase information, and if applicable the time information, are paired and associated with a unique positional information, in particular a unique directional information.

[0014] An object in the sense of the present invention is an area or reflection point on an object from which a radar signal can be reflected. An object may have one or more such objects. If an object has multiple objects, the radar signal may be reflected from them in different ways, for example in different directions. An object detected by a radar system may be referred to as a "detected object" for ease of distinction.

[0015] The radar system may determine at least directional information characterizing the direction of an object, and may be used to determine range and / or speed information characterizing the range and / or speed of an object relative to the radar system and / or the vehicle.

[0016] The intensity information, phase information, and position information, especially directional information, may comprise analog or digital signals, analog or digital data, real, complex or complex valued data, signals, numbers or values.

[0017] Depending on the means for signal processing, the received data may comprise electrical signals or electrical values ​​based on digital values, e.g., bits. In this way, the received data can be processed by the electrical means for signal processing. Additionally or alternatively, the received data may comprise optical signals or optical values ​​based on, e.g., quantum bits. In this way, the received data can be processed by the optical means for signal processing, e.g., a quantum processor.

[0018] Advantageously, the echo signals can be converted into received data by an analog-to-digital converter, so that the received data can be determined as digital data.

[0019] The reference frame for the radar system may include a reference point, a reference line, and / or a reference plane. If the radar system is used on a vehicle, the reference frame may also relate to the vehicle. In this case, the reference frame may include a reference point, a reference line, and / or a reference plane, such as a virtual axis of the vehicle, e.g., a longitudinal axis, a vertical axis, or a lateral axis of the vehicle. Advantageously, one reference frame may be a spherical coordinate system. The azimuth angle and the elevation angle may serve as position information characterizing the direction of the detected object. Additionally or alternatively, one reference frame may be a Cartesian coordinate system. The x, y, and z coordinates may serve as position information characterizing the position of the detected object. A vector having the x, y, and z coordinates may describe the direction of the detected object.

[0020] The present invention can be used in radar systems for vehicles, particularly automobiles. Advantageously, the present invention can be used in land vehicles, particularly cars, trucks, buses, motorcycles, etc., as well as aircraft and / or watercraft. The present invention is also applicable to radar systems for autonomous or partially autonomous vehicles. However, the present invention is not limited to vehicles. The present invention can also be applied to stationary vehicles, robotics, drones, and / or machinery, particularly construction or transport machinery such as cranes, excavators, etc.

[0021] The radar system may advantageously be connected to or be part of at least one control device, in particular a driver assistance system, of the vehicle or machine, thus enabling autonomous or partially autonomous operation of the vehicle or machine.

[0022] The present invention can be used in radar systems designed as a front radar system, a corner radar system, a rear radar system, a roof radar system, an underbody radar system, or an interior radar system of a vehicle. Thus, the radar system can be a front radar system, a corner radar system, a rear radar system, a roof radar system, an underbody radar system, or an interior radar system of a vehicle.

[0023] Radar systems can be used to detect objects, stationary or moving, in particular vehicles, people, animals, obstacles, road irregularities, in particular potholes and stones, road limits, open spaces, in particular parking spaces, precipitation, etc.

[0024] According to a preferred embodiment, as the at least one positional information at least one directional information, in particular an azimuth angle and / or an elevation angle and / or a vector, can be determined which characterizes the direction of the at least one detected object relative to the reference system.

[0025] Advantageously, the direction information may be an angle, in particular an angle of arrival (AoA), in this way the direction from which the echo signal arrives may be detected.

[0026] Advantageously, an azimuth angle and / or an elevation angle can be determined as the direction information. In this way, the position of the at least one object can be defined in a spherical coordinate system. Additionally or alternatively, a vector can be determined as the at least one position information. In this way, the position of the at least one object can be defined in a Cartesian coordinate system. Such a vector can include Cartesian coordinates, in particular x, y and / or z coordinates.

[0027] According to another preferred embodiment, at least one intensity information, at least one phase information, and, if applicable, time information can be realized as complex values, in particular as complex vectors. In this way, the intensity information, the phase information, and, if applicable, time information can be combined very efficiently. The intensity information, the phase information, and, if applicable, time information can then be efficiently processed by mathematical algorithms.

[0028] According to another preferred embodiment, additionally, time information can be determined for at least some of the virtual antenna elements. In this way, the pattern of the 2D FFT can also have a time component, since the phase relationship changes due to the movement of at least one object. The patterns of both the intensity information and the phase information can possibly have a time-dependent signature.

[0029] According to another preferred embodiment, at least a portion of the data of the array data set may be fed to at least one neural network designed as a spiking neural network.

[0030] In this way, the position information of moving objects that change their intensity and / or phase information over time can also be learned by at least one neural network. Spiking neural networks are well suited to the temporal content of typical radar echo signals, both in terms of time and processing parameters (angle, range, Doppler value, etc.). Spiking neural networks incorporate the concept of time into their structure. Spiking neural networks can reduce the complexity and power consumption of radar systems. Spiking neural networks can be efficiently implemented in neuromorphic hardware.

[0031] Alternatively, a neural processing unit (NPU) combined with a convolutional neural network (CNN) can be used. The neural processing unit can be hardware accelerated. Therefore, only the intensity and phase images can be used as the training data set for the neural network.

[0032] According to another preferred embodiment, an antenna system having transmitting and receiving antenna elements can be arranged and operated to form an undersampling system and / or an antenna system having transmitting and receiving antenna elements can be arranged and operated to form a sparse virtual antenna array. In this way, the ratio between the aperture and the number of required antenna elements is improved. The aperture of the virtual antenna array can be enlarged.

[0033] The virtual antenna elements are the sampling points of the echo signal. The extent of the virtual antenna array defines the aperture of the antenna configuration. The angular resolution relates to the physical distance between the most widely spaced virtual antenna elements relative to the wavelength of the radar signal.

[0034] Advantageously, the virtual antenna elements are arranged in one plane, in this way alignment of the virtual antenna array may be easier.

[0035] Advantageously, the four most distant virtual antenna elements can be arranged at the corners of a rectangle, in particular a square. In this way, a rectangular virtual antenna array can be realized. The aperture can therefore be clearly defined in two orthogonal directions, for example, in the azimuth and elevation directions. The four most distant virtual antenna elements can be arranged at the corners of a square. The apertures in the two orthogonal directions are therefore equal. In this way, the angular resolution in both orthogonal directions can be defined as a whole.

[0036] Advantageously, the length of the sides of the rectangle may in each case correspond to an integer multiple of half the wavelength of the radar signal, such that the distance between the virtual antenna elements arranged at adjacent corners of the rectangle each corresponds to an integer multiple of half the wavelength of the radar system.

[0037] By spacing the virtual antenna elements far enough apart, an undersampled system can be achieved, and the aperture will therefore exhibit spacing aliasing, which will obscure the determined position information, especially the angle of arrival.

[0038] According to another preferred embodiment, the four most distant virtual antenna elements may be arranged at the corners of a rectangle, in particular a square, with sides each corresponding to an integer multiple of half the wavelength of the radar signal. At least one additional virtual antenna element may be arranged at a distance of approximately half the wavelength of the radar signal from one of the four most distant virtual antenna elements. In this way, position information of at least one object may be determined with a reasonable aperture.

[0039] According to another preferred embodiment, a two-stage learning technique can be applied to the at least one neural network training. In this way, the at least one neural network can be trained very efficiently.

[0040] Advantageously, the two-stage learning technique can begin by feeding the neural network with an idealized artificial array data set. The idealized array data set can characterize an idealized aperture pattern. In this way, the network can learn the learning principles. Then, so-called starter enhancement can be performed. For this purpose, the complexity of the idealized artificial array data set can be increased, in particular by adding noise and / or random variations. Finally, the neural network can be fed with an actually recorded array data set.

[0041] According to another preferred embodiment, the at least one neural network may be trained on the array data set from both intensity and phase information. In this way, even complex array data sets characterizing complex aperture patterns may be recognized by the at least one neural network.

[0042] Furthermore, the object of the invention is achieved in a radar system in that the radar system comprises at least part of the means for carrying out the method according to the invention.

[0043] According to the invention, a radar system is designed as a MIMO radar, the radar system includes a plurality of transmitting antenna elements and a plurality of receiving antenna elements, and includes means for controlling the transmitting and receiving antenna elements such that during MIMO operation a virtual antenna array is generated with a plurality of virtual antenna elements for receiving echo signals.

[0044] Furthermore, the means for implementing the method according to the invention includes means for performing at least one two-dimensional fast Fourier transform to determine intensity and phase information from the received data.

[0045] Furthermore, the means for implementing the method according to the invention includes means for determining an array data set including at least magnitude and phase information of the virtual antenna elements.

[0046] Still further, the means for performing the method according to the invention comprise means for implementing at least one neural network to which data of the array data set can be fed in order to determine position information of the detected objects.

[0047] Advantageously, the means for implementing the method according to the invention may comprise at least one neural network, in particular at least one spiking neural network, in this way location information of the detected objects can be determined from the data of the array dataset.

[0048] At least part of the means for implementing the method according to the invention may be realized by software. In this way, in particular flow charts, in particular programs, algorithms and / or execution tables for implementing the method may be stored in the radar system. Additionally or alternatively, at least part of the means for implementing the method according to the invention may be realized by hardware.

[0049] Furthermore, the object of the invention is achieved in a driver assistance system in that the driver assistance system comprises at least part of the means for carrying out the method according to the invention.

[0050] According to the present invention, the driver assistance system comprises at least one radar system, in particular at least one radar system according to the present invention. Advantageously, the at least one radar system of the driver assistance system may comprise part of the means for carrying out the method according to the present invention. Since the at least one radar system is part of the driver assistance system, the means of the at least one radar system are also part of the driver assistance system. This also applies to the means of the vehicle having the at least one driver assistance system and / or the at least one radar system.

[0051] Furthermore, the object of the present invention is achieved in a vehicle in that the vehicle comprises at least part of the means for carrying out the method according to the present invention.

[0052] The vehicle is equipped with at least one radar system that may monitor the environment of the vehicle and / or the interior of the vehicle.

[0053] Advantageously, the vehicle may be equipped with at least one driver assistance system that allows the information obtained by the at least one radar system to be used for autonomous or at least partially autonomous driving of the vehicle.

[0054] Advantageously, the at least one radar system may be part of or connected to at least one driver assistance system, so that information acquired by the at least one radar system can be transmitted to a control unit of the at least one driver assistance system.

[0055] Additionally or alternatively, at least some of the means for performing the method according to the invention may be realized separately from the at least one radar system, for example together with a control unit of the vehicle and / or a control unit of a driver assistance assistance system.

[0056] In other respects, the features and advantages given with respect to the method according to the invention, the radar system according to the invention, the driver assistance system according to the invention, and the vehicle according to the invention, as well as their respective advantageous embodiments, apply mutatis mutandis, and vice versa. Naturally, the individual features and advantages can be combined with one another, which may result in further advantageous effects that exceed the sum of the individual effects.

[0057] The present invention, together with these and other objects and advantages, can be best understood from the following detailed description of the preferred embodiments, but is not limited to the embodiments. [Brief explanation of the drawings]

[0058] [Figure 1] FIG. 1 is a plan view of a vehicle having a driver assistance system with a radar system. [Figure 2] FIG. 2 is a side view of the vehicle of FIG. 1. [Figure 3] 3 is a functional diagram of a vehicle having the driver assistance system of FIGS. 1 and 2; [Figure 4]4 shows a virtual antenna array created by the transmit and receive antenna arrays of the vehicle radar system of FIGS. 1 to 3 in a MIMO operating mode according to the first embodiment; [Figure 5] 4 shows a virtual antenna array created by the transmit and receive antenna arrays of the vehicle radar system of FIGS. 1 to 3 in a MIMO operating mode according to a second embodiment; [Figure 6] 4 shows a virtual antenna array created by the transmit and receive antenna arrays of the vehicle radar system of FIGS. 1 to 3 in a MIMO operating mode according to a third embodiment; [Figure 7] 4 is a flowchart of a method of operating the radar system of the vehicle of FIGS. 1-3. [Figure 8] Intensity array image visualizing the aperture pattern of intensity information for multiple targets detected by the radar system of the vehicle in Figures 1 to 3. DETAILED DESCRIPTION OF THE INVENTION

[0059] In the drawings, identical or similar elements are marked with the same reference numerals. The drawings are merely schematic representations and are not intended to portray specific parameters of the invention. Moreover, the drawings depict only typical embodiments of the invention and should not be considered as limiting the scope of the invention.

[0060] Figure 1 shows a plan view of a vehicle 10 in the form of a passenger car. Figure 2 shows the vehicle 10 in a side view.

[0061] The vehicle 10 is equipped with a driver assistance system 12. Figure 3 shows a functional diagram of the vehicle 10 with the driver assistance system 12. The driver assistance system 12 enables the vehicle 10 to operate partially autonomously or autonomously.

[0062] The driver assistance system 12 includes a radar system 14 and a control unit 16. The radar system 14 allows monitoring of the environment ahead of the vehicle 10. The radar system 14 is connected to the control unit 16 so that data collected by the radar system 14 about the environment can be transmitted to the control unit 16. The control unit 16 of the driver assistance system 12 allows driving functions of the vehicle 10 to be controlled based on information obtained by the radar system 14.

[0063] The radar system 14 is exemplarily arranged in a forward region of the vehicle 10, for example on the front bumper. The radar system 14 can be used to monitor a surveillance area ahead of the vehicle 10 in the direction of travel, for example for an object 18. In Figures 1 to 3, an object 18 is shown as an example. The radar system 14 may be arranged at different positions on the vehicle 10 and may be oriented in different directions. Multiple radar systems 14 may be provided.

[0064] The radar system 14 can detect objects 20, such as stationary or moving objects 18, including vehicles, people, animals, plants, obstacles, the ground, roads, road irregularities such as potholes and stones, road boundaries, (traffic) signs, traffic lights, open spaces such as parking spaces, precipitation, etc.

[0065] An object 20 in the sense of the present invention is an area or reflection point of the object 18 from which a radar signal 22 can be reflected. The object 18 may have one or more such objects 20. If the object 18 has multiple objects 20, the radar signal 22 may be reflected from them in different ways, for example in different directions. The objects 20 detected by the radar system 14 may be referred to as detected objects 20 for ease of distinction. For the sake of clarity, only two objects 20 of the object 18 are shown by way of example in FIGS. 1 to 3 .

[0066] The radar system 14 may determine the distance 24, direction, and velocity of the object 18 relative to the reference frame of the vehicle 10. The reference frame may be, for example, a spherical coordinate system. The azimuth angle Φ and the elevation angle Θ serve as directional information that characterize the direction of the detected object 20. The origin of the spherical coordinate system is located at the intersection of the longitudinal axis 26 of the vehicle 10 and the vertical axis 28 of the vehicle 10. The azimuth angle Φ=0° is on the longitudinal axis 26 of the vehicle 10.

[0067] The radar system 14 is designed as a MIMO (multiple-input multiple-output) radar. The radar system 14 includes a control and evaluation device 30, a transmitting antenna array 32 having a plurality of transmitting antenna elements 34 for transmitting electromagnetic radar signals 22, and a receiving antenna array 36 having a plurality of receiving antenna elements 38 for receiving electromagnetic echo signals 40. By way of example, two transmitting antenna elements 34 and two receiving antenna elements 38 are shown in FIG. 3.

[0068] During MIMO operation of the radar system 14, the transmit antenna elements 34 and the receive antenna elements 38 create a virtual antenna array 42 having a plurality of virtual antenna elements 44 for receiving the echo signals 40. The virtual antenna array 42 is implemented as a sparse virtual antenna array.

[0069] FIG. 4 shows a first example of a virtual antenna array 42. The virtual antenna array 42 includes six virtual antenna elements 44. The virtual antenna elements 44 are arranged on a plane. The four most distant virtual antenna elements 44 are arranged at the corners of a square. In this manner, a square virtual antenna array 42 is realized. The horizontal extent of the virtual antenna array 42 defines a horizontal aperture 46 of the antenna array, which includes the transmit antenna array 32 and the receive antenna array 36. The vertical extent defines a vertical aperture 48 of the antenna array. The spacing between the virtual antenna elements 44 arranged at adjacent corners of the square corresponds to an integer multiple of half the wavelength of the radar system 14. The fifth virtual antenna element 44 is arranged on the connecting line between the virtual antenna element 44 at the lower left corner of the square and the virtual antenna element 44 at the upper left corner, at a spacing 50 of approximately half the wavelength of the radar signal 22 from the virtual antenna element 44 at the lower left corner. The sixth virtual antenna element 44 is located on the connecting line between the virtual antenna element 44 in the lower left corner and the virtual antenna element 44 in the lower right corner, at a distance 52 of approximately half the wavelength of the radar signal 22 from the virtual antenna element 44 in the lower left corner.

[0070] Figure 5 shows a second example of the virtual antenna array 42. Unlike the virtual antenna array 42 shown in Figure 4, in the second example, the sixth virtual antenna element 44 is arranged on the connecting line between the virtual antenna element 44 in the upper left corner and the virtual antenna element 44 in the upper right corner, at a distance 54 of approximately half the wavelength of the radar signal 22 from the virtual antenna element 44 in the upper left corner.

[0071] Figure 6 shows a third example of a virtual antenna array 42. Unlike the virtual antenna array 42 shown in Figure 4, the third example includes only five virtual antenna elements 44. Four of the virtual antenna elements 44 are located at the corners of a square, as in the first example of Figure 4. The fifth virtual antenna element 33 is located on the connecting line between the virtual antenna element 44 in the lower left corner and the virtual antenna element 44 in the upper right corner, at a distance 56 of approximately half the wavelength of the radar signal 22 from the virtual antenna element 44 in the lower left corner.

[0072] Furthermore, the control and evaluation device 30 includes means, for example an analog-to-digital converter 58, for converting the electromagnetic echo signals 40 into received data 72 suitable for signal processing.

[0073] The control and evaluation unit 30 includes a Fourier transform means 60 that performs a two-dimensional fast Fourier transform 74 to determine magnitude and phase information from the received data 72. In combination with the sparse virtual antenna array 42, an under-sampled system may be formed.

[0074] Additionally, the control and evaluation unit 30 includes a data set means 62 for determining an array data set 80 including magnitude and phase information acquired by the virtual antenna elements 44 .

[0075] Furthermore, the control and evaluation unit 30 includes a neural network 64. Data from the array data set 80 can be fed to the neural network 64 to determine position information characterizing the orientation of the detected object 20. The neural network 64 is designed as a spiking neural network 64. The spiking neural network 64 is well suited to the temporal information content of the echo signal 40.

[0076] The control and evaluation device 30 also includes a storage medium 66 on which the data of the array data set 80, the position information, and the training data for the neural network 64 may be stored.

[0077] At least a portion of the means for implementing the method for operating the radar system 14 may be realized by software. The storage medium 66 of the control and evaluation device 30 may store, for example, a flowchart, e.g., a program, an algorithm, and / or an execution table for implementing the method.

[0078] A method of operating the radar system 14 is described in more detail below with reference to the flow chart of FIG.

[0079] A series of measurements involving multiple radar measurements is performed by the radar system 14 .

[0080] In process step 68, radar signals 22 are transmitted by each transmit antenna element 34 for each measurement according to the MIMO mode of operation. During the MIMO mode of operation, the transmit antenna elements 34 and the receive antenna elements 38 form a virtual antenna array 42 having a plurality of antenna elements 44. When an object 18 is present within the field of view of the radar system 14, the radar signals 22 reflect off objects 20 on the object 18. Electromagnetic echo signals 40 resulting from the radar signals 22 reflected from the objects 20 are received by the virtual antenna elements 44.

[0081] In a processing step 70, the received echo signals 40 are converted by an analog-to-digital converter 58 into received data 72 suitable for further signal processing.

[0082] Each antenna element of the virtual antenna array 44 contains complex values ​​representing the magnitude and phase information corresponding to the echo signals within a single range-Doppler cell, as well as information regarding the number of targets sampled by the receive array and their orientation (azimuth and elevation). This complex information is then transformed using a two-dimensional Fourier transform 74. The resulting aperture pattern (e.g., FIG. 8) is presented as magnitude and phase information and may also be realized as complex magnitude / phase values.

[0083] In processing step 78, an array data set 80 is determined for each measurement. The array data set 80 includes complex intensity / phase values ​​76 containing intensity and phase information for all virtual antenna elements 44. Additionally, the array data set 80 includes time information for each measurement. The time information characterizes the time position of the measurement in a series of measurements within a single range-Doppler cell. The time information can be used to characterize the movement of objects 20 relative to each other and can be used to separate multiple objects in the same range-Doppler bin.

[0084] 8 shows an intensity array image as a grayscale representation visualizing the aperture pattern of intensity information of an exemplary array data set 80 for multiple detected targets 20 versus the angle of arrival (AoA) of the received echo signals 40. The angle of arrival is characterized by an azimuth angle Φ and an elevation angle Θ. The intensity information is defined by a linear grayscale shown adjacent to the intensity array image.

[0085] A phase array image (not shown) as a grayscale display visualizing an aperture pattern of phase information for the angle of arrival of the received echo signals 40 can be generated according to the intensity array image of FIG.

[0086] After determining the array data set 80, the data in the array data set 80 is fed to a neural network 64, which can operate on both scalar or complex values. The neural network 64 can determine position information in the form of an azimuth angle Φ and an elevation angle Θ for each detected object 20. The intensity, phase, and time information are treated as a single image from which directional information is learned using the neural network 64. The intensity and phase images are paired with corresponding time information and associated with unique directional information.

[0087] The azimuth angle Φ and elevation angle Θ of each detected object 20 are transmitted to the control unit 16 of the driver assistance system 12. The driver assistance system 12 controls the driving functions of the vehicle 10 based on the position information of the detected objects 20.

[0088] Optionally, prior to normal operation of the radar system 14, a two-stage training technique may be implemented to train the neural network 64. The two-stage training technique may begin by providing the neural network 64 with an idealized artificial array data set 80. The idealized array data set 80 characterizes an idealized aperture pattern from both intensity and phase information. In this way, the neural network 64 can learn the learning principles. Then, so-called starter strengthening may be implemented. For this purpose, the complexity of the idealized artificial array data set 80 may be increased, for example, by adding noise and / or random fluctuations. Finally, the neural network 64 may be provided with an actually recorded array data set 80.

Claims

1. A method of operating a radar system (14), particularly a radar system (14) for a vehicle (10), comprising: At least one electromagnetic radar signal (22) is transmitted from at least one transmit antenna element (34) of the radar system (14); At least one electromagnetic echo signal (40) resulting from at least one radar signal (22) reflected from at least one object (20) within the field of view of the radar system (14) is received by at least one receive antenna element (38) and converted into receive data (72) suitable for signal processing; at least one magnitude information (76) and at least one phase information (76) corresponding to at least one received echo signal (40) is determined from at least a portion of the received data (72); 1. A method for detecting a position of at least one object detected by a radar system, comprising: determining at least one position information (Φ; Θ) using machine learning (64), the at least one position information (Φ; Θ) characterizing at least an orientation of at least one detected object relative to a frame of reference for the radar system (14), the method comprising: the radar system (14) is operated as a MIMO radar including a plurality of transmit antenna elements (34) and a plurality of receive antenna elements (38), and during MIMO operation of the radar system (14), the transmit antenna elements (34) and the receive antenna elements (38) generate a virtual antenna array (42) having a plurality of virtual antenna elements (44) for receiving echo signals (40); performing at least one two-dimensional fast Fourier transform (74) on at least a portion of the virtual antenna elements (44) to determine each of magnitude information (76) and phase information (76) from at least a portion of the received data (72); determining at least one array data set (80) including at least the magnitude information (76) and the phase information (76) for at least the portion of the virtual antenna elements (44); At least a portion of the data of at least one of said array data sets (80) is fed to at least one neural network (64) to determine at least one position information (Φ; Θ) of at least one of said detected objects (20); A method characterized by:

2. as the at least one positional information, at least one directional information characterizing the orientation of the at least one detected object (20) relative to the reference system is determined, in particular an angle such as an azimuth angle (Φ) and / or an elevation angle (Θ) and / or a vector; 2. The method of claim 1 .

3. said at least one intensity information and said at least one phase information and, if applicable, time information, are realized as complex values ​​(76), in particular as complex vectors; 3. The method according to claim 1 or 2.

4. Additionally, time information is determined for at least some of the virtual antenna elements (44).

4. The method according to claim 1, wherein the first and second electrodes are connected to a first electrode.

5. At least a portion of the data of the array data set (80) is fed to at least one neural network (64) designed as a spiking neural network.

5. The method according to claim 1, wherein the first and second electrodes are connected to a first electrode.

6. an antenna system (32, 36) having the transmitting antenna elements (34) and the receiving antenna elements (38) arranged and operated to form an undersampling system, and / or an antenna array (32, 36) having the transmitting antenna elements (34) and the receiving antenna elements (38) arranged and operated to form a sparse virtual antenna array (42); 6. The method according to any one of claims 1 to 5.

7. the four most distant virtual antenna elements (44) are arranged at the corners of a rectangle, in particular a square, the lengths (46, 48) of the sides of which correspond to an integer multiple of half the wavelength of the radar signal (22), and at least one additional virtual antenna element (44) is arranged at a distance (50, 52; 54; 56) of approximately half the wavelength of the radar signal (22) from one of the four most distant virtual antenna elements (44); 7. The method according to any one of claims 1 to 6.

8. a two-stage training technique is applied to train at least one said neural network (64); 8. The method according to any one of claims 1 to 7.

9. At least one of the neural networks (64) is trained on an array data set (80) from both the intensity information (76) and the phase information (76).

9. The method according to any one of claims 1 to 8.

10. at least one transmitting antenna element (34) for transmitting an electromagnetic radar signal (22); at least one receiving antenna element (38) for receiving an electromagnetic echo signal (40); means for converting the electromagnetic echo signals (40) into received data (72) suitable for signal processing; means for determining magnitude information (76) and phase information (76) from the received data (72); means for determining, using machine learning (64), at least one positional information (Φ; Θ) characterizing at least said orientation of at least one detected object (20); A radar system (14), in particular a radar system (14) for a vehicle (10), comprising: The radar system (14) comprises at least part of the means for carrying out the method according to any one of claims 1 to 9. A radar system (14) characterized in that

11. A driver assistance system (12) comprising at least one radar system (14), said at least one radar system (14) comprising: at least one transmitting antenna element (34) for transmitting an electromagnetic radar signal (22); at least one receiving antenna element (38) for receiving an electromagnetic echo signal (40); means for converting the electromagnetic echo signals (40) into received data (72) suitable for signal processing; means for determining magnitude information (76) and phase information (76) from the received data (72); means for determining, using machine learning (64), at least one positional information (Φ; Θ) characterizing at least said orientation of at least one detected object (20); A driver assistance system (12) comprising: The driver assistance system (12) comprises at least part of the means for implementing the method according to any one of claims 1 to 9. A driver assistance system (12) characterized in that

12. A vehicle (10) comprising at least one radar system (14), the at least one radar system (14) comprising: at least one transmitting antenna element (34) for transmitting an electromagnetic radar signal (22); at least one receiving antenna element (38) for receiving an electromagnetic echo signal (40); means for converting the electromagnetic echo signals (40) into received data (72) suitable for signal processing; means for determining magnitude information (76) and phase information (76) from the received data (72); means for determining, using machine learning (64), at least one positional information (Φ; Θ) characterizing at least said orientation of at least one detected object (20); In a vehicle (10) equipped with The vehicle (10) comprises at least part of the means for carrying out the method according to any one of claims 1 to 9. A vehicle (10) characterized in that:

Images